Gis-Based Landslide Susceptibility Modeling Using The Weights Of Evidence And Logistic Regression Methods In Bichena And Yed Wiha Area, East Gojam, Central Ethiopia

Abstract

Landslide Distribution And Susceptibility Mapping Are Fundamental Steps For Landslide Related Hazard And Risk Management Activities. This Is Especially Crucial In A Country Like Ethiopia. It Has Mountainous Terrain, Heterogeneous Rock Units With Varying Degrees Of Weathering, And Complex Hydrology That Contribute To Landslide Initiation. Landslides Can Result In Loss Of Life, Property Damage, And Infrastructure Disruption. The Primary Purpose Of This Study Is To Prepare A Landslide Susceptibility Map For The Bichena And Yed Wiha Area Using The Weight Of Evidence (Woe) And Logistic Regression (Lr) Methods That Use Continuous And Discrete Variables Efficiently. Data Were Collected From Different Sources To Delineate Landslide Susceptibility Zones In The Study Area To Produce Better Results Of A Landslide Susceptibility Map. Eight Triggering Factors Were Considered Namely; Aspect, Slope, Curvature, Distance From Stream, Distance From Lineament, Lithology, Land Use, And Rainfall. Rainfall-Induced Landslides Of Different Types Frequently Affect The Hilly And Mountainous Terrains Of The Highlands Of Ethiopia And Are Weighted Using The Weight Of Evidence. The Weight Of Each Factor Will Be Calculated And Assigned In Aeronautical Reconnaissance Coverage Geographic Information System (Arc-Gis). To Add These Factors In Arc-Gis Spatial Analyst Tools Raster Calculator, And Produce A Landslide Susceptibility Map, The Weighted Contrast Linear Combination And Coefficient Of Regression Were Used Between Training Landslide And Landslide Causative Factors. To Establish The Relational Statistical Correlation Of Various Causative Factors With Past Landslides In The Area, Landslide Inventory Were Made Through Google Earth Image Interpretation. Further, The Weight Of Evidence And Logistic Regression, Bivariate, And Multivariate Statistical Analysis Methods Were Used To Prepare The Model. Statistical Package For The Social Science (Spss) Software And Ms Excel Were Used To Check The Validation Of The Overlay Maps Using Receiver Operating Characteristic (Roc) And Landslide Density Index (Ldi). Finally, A Landslide Susceptibility Map Was Prepared And Classified Into Five Classes: Very Low, Low, Medium, High, And Very High. Finally, Landslide Mitigation Measures Were Recommended For Susceptible Areas.